Towards a New Understanding of the Training of Neural Networks with Mislabeled Training Data
We investigate the problem of machine learning with mislabeled training data. We try to make the effects of mislabeled training better understood through analysis of the basic model and equations that characterize the problem. This includes results about the ability of the noisy model to make the same decisions as the clean model and the effects of noise on model performance. In addition to providing better insights we also are able to show that the Maximum Likelihood (ML) estimate of the parameters of the noisy model determine those of the clean model. This property is obtained through the use of the ML invariance property and leads to an approach to developing a classifier when training has been mislabeled: namely train the classifier on noisy data and adjust the decision threshold based on the noise levels and/or class priors. We show how our approach to mislabeled training works with multi-layered perceptrons (MLPs).
Causal Modeling for Fairness in Dynamical Systems
In this work, we present causal directed acyclic graphs (DAGs) as a unifying framework for the recent literature on fairness in dynamical systems. We advocate for the use of causal DAGs as a tool in both designing equitable policies and estimating their impacts. By visualizing models of dynamic unfairness graphically, we expose implicit causal assumptions which can then be more easily interpreted and scrutinized by domain experts. We demonstrate that this method of reinterpretation can be used to critique the robustness of an existing model/policy, or uncover new policy evaluation questions. Causal models also enable a rich set of options for evaluating a new candidate policy without incurring the risk of implementing the policy in the real world. We close the paper with causal analyses of several models from the recent literature, and provide an in-depth case study to demonstrate the utility of causal DAGs for modeling fairness in dynamical systems.
CASA-NLU: Context-Aware Self-Attentive Natural Language Understanding for Task-Oriented Chatbots
Natural Language Understanding (NLU) is a core component of dialog systems. It typically involves two tasks – intent classification (IC) and slot labeling (SL), which are then followed by a dialogue management (DM) component. Such NLU systems cater to utterances in isolation, thus pushing the problem of context management to DM. However, contextual information is critical to the correct prediction of intents and slots in a conversation. Prior work on contextual NLU has been limited in terms of the types of contextual signals used and the understanding of their impact on the model. In this work, we propose a context-aware self-attentive NLU (CASA-NLU) model that uses multiple signals, such as previous intents, slots, dialog acts and utterances over a variable context window, in addition to the current user utterance. CASA-NLU outperforms a recurrent contextual NLU baseline on two conversational datasets, yielding a gain of up to 7% on the IC task for one of the datasets. Moreover, a non-contextual variant of CASA-NLU achieves state-of-the-art performance for IC task on standard public datasets – Snips and ATIS.
Detailed comparison of communication efficiency of split learning and federated learning
We compare communication efficiencies of two compelling distributed machine learning approaches of split learning and federated learning. We show useful settings under which each method outperforms the other in terms of communication efficiency. We consider various practical scenarios of distributed learning setup and juxtapose the two methods under various real-life scenarios. We consider settings of small and large number of clients as well as small models (1M – 6M parameters), large models (10M – 200M parameters) and very large models (1 Billion-100 Billion parameters). We show that increasing number of clients or increasing model size favors split learning setup over the federated while increasing the number of data samples while keeping the number of clients or model size low makes federated learning more communication efficient.
Multivariate Rank-based Distribution-free Nonparametric Testing using Measure Transportation
In this paper, we propose a general framework for distribution-free nonparametric testing in multi-dimensions, based on a notion of multivariate ranks defined using the theory of measure transportation. Unlike other existing proposals in the literature, these multivariate ranks share a number of useful properties with the usual one-dimensional ranks; most importantly, these ranks are distribution-free. This crucial observation allows us to design nonparametric tests that are exactly distribution-free under the null hypothesis. We demonstrate the applicability of this approach by constructing exact distribution-free tests for two classical nonparametric problems: (i) testing for mutual independence between random vectors, and (ii) testing for the equality of multivariate distributions. In particular, we propose (multivariate) rank versions of distance covariance (Sz\’ekely et al., 2007) and energy statistic (Sz\’ekely and Rizzo, 2013) for testing scenarios (i) and (ii) respectively. In both these problems, we derive the asymptotic null distribution of the proposed test statistics. We further show that our tests are consistent against all fixed alternatives. Moreover, the proposed tests are tuning-free, computationally feasible and are well-defined under minimal assumptions on the underlying distributions (e.g., they do not need any moment assumptions). We also demonstrate the efficacy of these procedures via extensive simulations. In the process of analyzing the theoretical properties of our procedures, we end up proving some new results in the theory of measure transportation and in the limit theory of permutation statistics using Stein’s method for exchangeable pairs, which may be of independent interest.
BPMR: Bayesian Probabilistic Multivariate Ranking
Multi-aspect user preferences are attracting wider attention in recommender systems, as they enable more detailed understanding of users’ evaluations of items. Previous studies show that incorporating multi-aspect preferences can greatly improve the performance and explainability of recommendation. However, as recommendation is essentially a ranking problem, there is no principled solution for ranking multiple aspects collectively to enhance the recommendation. In this work, we derive a multi-aspect ranking criterion. To maintain the dependency among different aspects, we propose to use a vectorized representation of multi-aspect ratings and develop a probabilistic multivariate tensor factorization framework (PMTF). The framework naturally leads to a probabilistic multi-aspect ranking criterion, which generalizes the single-aspect ranking to a multivariate fashion. Experiment results on a large multi-aspect review rating dataset confirmed the effectiveness of our solution.
Bayesian Analysis of Multidimensional Functional Data
Multi-dimensional functional data arises in numerous modern scientific experimental and observational studies. In this paper we focus on longitudinal functional data, a structured form of multidimensional functional data. Operating within a longitudinal functional framework we aim to capture low dimensional interpretable features. We propose a computationally efficient nonparametric Bayesian method to simultaneously smooth observed data, estimate conditional functional means and functional covariance surfaces. Statistical inference is based on Monte Carlo samples from the posterior measure through adaptive blocked Gibbs sampling. Several operative characteristics associated with the proposed modeling framework are assessed comparatively in a simulated environment. We illustrate the application of our work in two case studies. The first case study involves age-specific fertility collected over time for various countries. The second case study is an implicit learning experiment in children with Autism Spectrum Disorder (ASD).
A literature review on current approaches and applications of fuzzy expert systems
The main purposes of this study are to distinguish the trends of research in publication exits for the utilisations of the fuzzy expert and knowledge-based systems that is done based on the classification of studies in the last decade. The present investigation covers 60 articles from related scholastic journals, International conference proceedings and some major literature review papers. Our outcomes reveal an upward trend in the up-to-date publications number, that is evidence of growing notoriety on the various applications of fuzzy expert systems. This raise in the reports is mainly in the medical neuro-fuzzy and fuzzy expert systems. Moreover, another most critical observation is that many modern industrial applications are extended, employing knowledge-based systems by extracting the experts’ knowledge.
Absum: Simple Regularization Method for Reducing Structural Sensitivity of Convolutional Neural Networks
We propose Absum, which is a regularization method for improving adversarial robustness of convolutional neural networks (CNNs). Although CNNs can accurately recognize images, recent studies have shown that the convolution operations in CNNs commonly have structural sensitivity to specific noise composed of Fourier basis functions. By exploiting this sensitivity, they proposed a simple black-box adversarial attack: Single Fourier attack. To reduce structural sensitivity, we can use regularization of convolution filter weights since the sensitivity of linear transform can be assessed by the norm of the weights. However, standard regularization methods can prevent minimization of the loss function because they impose a tight constraint for obtaining high robustness. To solve this problem, Absum imposes a loose constraint; it penalizes the absolute values of the summation of the parameters in the convolution layers. Absum can improve robustness against single Fourier attack while being as simple and efficient as standard regularization methods (e.g., weight decay and L1 regularization). Our experiments demonstrate that Absum improves robustness against single Fourier attack more than standard regularization methods. Furthermore, we reveal that robust CNNs with Absum are more robust against transferred attacks due to decreasing the common sensitivity and against high-frequency noise than standard regularization methods. We also reveal that Absum can improve robustness against gradient-based attacks (projected gradient descent) when used with adversarial training.
Differentially Private Regression and Classification with Sparse Gaussian Processes
A continuing challenge for machine learning is providing methods to perform computation on data while ensuring the data remains private. In this paper we build on the provable privacy guarantees of differential privacy which has been combined with Gaussian processes through the previously published \emph{cloaking method}. In this paper we solve several shortcomings of this method, starting with the problem of predictions in regions with low data density. We experiment with the use of inducing points to provide a sparse approximation and show that these can provide robust differential privacy in outlier areas and at higher dimensions. We then look at classification, and modify the Laplace approximation approach to provide differentially private predictions. We then combine this with the sparse approximation and demonstrate the capability to perform classification in high dimensions. We finally explore the issue of hyperparameter selection and develop a method for their private selection. This paper and associated libraries provide a robust toolkit for combining differential privacy and GPs in a practical manner.
Shapes of Uncertainty in Spectral Graph Theory
We present a flexible framework for uncertainty principles in spectral graph theory. In this framework, general filter functions modeling the spatial and spectral localization of a graph signal can be incorporated. It merges several existing uncertainty relations on graphs, among others the Landau-Pollak principle describing the joint admissibility region of two projection operators, and uncertainty relations based on spectral and spatial spreads. Using theoretical and computational aspects of the numerical range of matrices, we are able to characterize and illustrate the shapes of the uncertainty curves and to study the space-frequency localization of signals inside the admissibility regions.
Exploring ways to incorporate additional knowledge to improve Natural Language Commonsense Question Answering
DARPA and Allen AI have proposed a collection of datasets to encourage research in Question Answering domains where (commonsense) knowledge is expected to play an important role. Recent language models such as BERT and GPT that have been pre-trained on Wikipedia articles and books, have shown decent performance with little fine-tuning on several such Multiple Choice Question-Answering (MCQ) datasets. Our goal in this work is to develop methods to incorporate additional (commonsense) knowledge into language model based approaches for better question answering in such domains. In this work we first identify external knowledge sources, and show that the performance further improves when a set of facts retrieved through IR is prepended to each MCQ question during both training and test phase. We then explore if the performance can be further improved by providing task specific knowledge in different manners or by employing different strategies for using the available knowledge. We present three different modes of passing knowledge and five different models of using knowledge including the standard BERT MCQ model. We also propose a novel architecture to deal with situations where information to answer the MCQ question is scattered over multiple knowledge sentences. We take 200 predictions from each of our best models and analyze how often the given knowledge is useful, how many times the given knowledge is useful but system failed to use it and some other metrices to see the scope of further improvements.
Data Augmentation Revisited: Rethinking the Distribution Gap between Clean and Augmented Data
Data augmentation has been widely applied as an effective methodology to prevent over-fitting in particular when training very deep neural networks. The essential benefit comes from introducing additional priors in visual invariance, and thus generate images in different appearances but containing the same semantics. Recently, researchers proposed a few powerful data augmentation techniques which indeed improved accuracy, yet we notice that these methods have also caused a considerable gap between clean and augmented data. This paper revisits this problem from an analytical perspective, for which we estimate the upper-bound of testing loss using two terms, named empirical risk and generalization error, respectively. Data augmentation significantly reduces the generalization error, but meanwhile leads to a larger empirical risk, which can be alleviated by a simple algorithm, i.e. using less-augmented data to refine the model trained on fully-augmented data. We validate our approach on a few popular image classification datasets including CIFAR and ImageNet, and demonstrate consistent accuracy gain. We also conjecture that this simple strategy implies a generalized approach to circumvent local minima, which is of value to future research on model optimization.
Extracting Conceptual Knowledge from Natural Language Text Using Maximum Likelihood Principle
Domain-specific knowledge graphs constructed from natural language text are ubiquitous in today’s world. In many such scenarios the base text, from which the knowledge graph is constructed, concerns itself with practical, on-hand, actual or ground-reality information about the domain. Product documentation in software engineering domain are one example of such base texts. Other examples include blogs and texts related to digital artifacts, reports on emerging markets and business models, patient medical records, etc. Though the above sources contain a wealth of knowledge about their respective domains, the conceptual knowledge on which they are based is often missing or unclear. Access to this conceptual knowledge can enormously increase the utility of available data and assist in several tasks such as knowledge graph completion, grounding, querying, etc. Our contributions in this paper are twofold. First, we propose a novel Markovian stochastic model for document generation from conceptual knowledge. The uniqueness of our approach lies in the fact that the conceptual knowledge in the writer’s mind forms a component of the parameter set of our stochastic model. Secondly, we solve the inverse problem of learning the best conceptual knowledge from a given document, by finding model parameters which maximize the likelihood of generating the specific document over all possible parameter values. This likelihood maximization is done using an application of Baum-Welch algorithm, which is a known special case of Expectation-Maximization (EM) algorithm. We run our conceptualization algorithm on several well-known natural language sources and obtain very encouraging results. The results of our extensive experiments concur with the hypothesis that the information contained in these sources has a well-defined and rigorous underlying conceptual structure, which can be discovered using our method.
Human-In-The-Loop Learning of Qualitative Preference Models
In this work, we present a novel human-in-the-loop framework to help the human user understand the decision making process that involves choosing preferred options. We focus on qualitative preference models over alternatives from combinatorial domains. This framework is interactive: the user provides her behavioral data to the framework, and the framework explains the learned model to the user. It is iterative: the framework collects feedback on the learned model from the user and tries to improve it accordingly till the user terminates the iteration. In order to communicate the learned preference model to the user, we develop visualization of intuitive and explainable graphic models, such as lexicographic preference trees and forests, and conditional preference networks. To this end, we discuss key aspects of our framework for lexicographic preference models.
Highlighting Bias with Explainable Neural-Symbolic Visual Reasoning
Many high-performance models suffer from a lack of interpretability. There has been an increasing influx of work on explainable artificial intelligence (XAI) in order to disentangle what is meant and expected by XAI. Nevertheless, there is no general consensus on how to produce and judge explanations. In this paper, we discuss why techniques integrating connectionist and symbolic paradigms are the most efficient solutions to produce explanations for non-technical users and we propose a reasoning model, based on definitions by Doran et al. [2017] (arXiv:1710.00794) to explain a neural network’s decision. We use this explanation in order to correct bias in the network’s decision rationale. We accompany this model with an example of its potential use, based on the image captioning method in Burns et al. [2018] (arXiv:1803.09797).
Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML
An important research direction in machine learning has centered around developing meta-learning algorithms to tackle few-shot learning. An especially successful algorithm has been Model Agnostic Meta-Learning (MAML), a method that consists of two optimization loops, with the outer loop finding a meta-initialization, from which the inner loop can efficiently learn new tasks. Despite MAML’s popularity, a fundamental open question remains — is the effectiveness of MAML due to the meta-initialization being primed for rapid learning (large, efficient changes in the representations) or due to feature reuse, with the meta initialization already containing high quality features? We investigate this question, via ablation studies and analysis of the latent representations, finding that feature reuse is the dominant factor. This leads to the ANIL (Almost No Inner Loop) algorithm, a simplification of MAML where we remove the inner loop for all but the (task-specific) head of a MAML-trained network. ANIL matches MAML’s performance on benchmark few-shot image classification and RL and offers computational improvements over MAML. We further study the precise contributions of the head and body of the network, showing that performance on the test tasks is entirely determined by the quality of the learned features, and we can remove even the head of the network (the NIL algorithm). We conclude with a discussion of the rapid learning vs feature reuse question for meta-learning algorithms more broadly.
Real-Time Verification for Distributed Cyber-Physical Systems
Safety-critical distributed cyber-physical systems (CPSs) have been found in a wide range of applications. Notably, they have displayed a great deal of utility in intelligent transportation, where autonomous vehicles communicate and cooperate with each other via a high-speed communication network. Such systems require an ability to identify maneuvers in real-time that cause dangerous circumstances and ensure the implementation always meets safety-critical requirements. In this paper, we propose a real-time decentralized reachability approach for safety verification of a distributed multi-agent CPS with the underlying assumption that all agents are time-synchronized with a low degree of error. In the proposed approach, each agent periodically computes its local reachable set and exchanges this reachable set with the other agents with the goal of verifying the system safety. Our method, implemented in Java, takes advantages of the timing information and the reachable set information that are available in the exchanged messages to reason about the safety of the whole system in a decentralized manner. Any particular agent can also perform local safety verification tasks based on their local clocks by analyzing the messages it receives. We applied the proposed method to verify, in real-time, the safety properties of a group of quadcopters performing a distributed search mission.
